Bedaquiline - New anti-tuberculosis treatment

III. New anti-tuberculosis treatment

11. Bedaquiline

Structure and mechanism of action

Bedaquiline represents the first compound of a novel class of anti-mycobacterial drugs, the diarylquinolines. Bedaquiline was previously known as TMC207 and R207910 during its development and is currently marketed with the commercial name of Sirturo^® by Janssen Pharmaceuticals (Belgium). Bedaquiline is the most promising among the diarylquinolines that have been identified through systematic testing of approximately 70 000 compounds on strains of M. smegmatis, which represents an easier target for drug development than M. tuberculosis.¹³¹ The conformation of bedaquiline has been characterised combining magnetic resonance imaging with molecular modelling,¹³² and subsequently confirmed using X-ray diffraction.¹³³ These studies allowed to define the specific structure of the quinoline nucleus of the diarylquinolines, characterised by the functionalized lateral (3’) chain, which differentiates them from other molecules containing a quinoline group.¹²² Figure 9 shows the structure and absolute conformation of bedaquiline, defined chemically as 1-(6-bromo-2-methoxy-quinolin-3-yl)-4-dimethylamino-2-naphthalen-1-yl-1-phenyl-butan-2-ol, or with the general formula C32H31BrN2O2; its molecular weight is 555.51 Da.¹³¹ The molecular target of bedaquiline is an ATP synthase, a very well-conserved membrane protein that plays a fundamental role in the energetic metabolism and is ubiquitarian across prokaryotes and eukaryotes. The mycobacterial ATP synthase is composed by a transmembrane and a cytoplasmic part, each constituted by different subunits (Figure 10). It is essential for the survival of mycobacteria, regardless of their metabolic status and environment: actively replicating and dormant, intra- and extracellular.¹³⁴ Bedaquiline affinity for the mycobacterial ATP synthase is much higher than the one for human ATP synthase, a reassuring finding in view of potential toxicity concerns.¹³⁵ The mutations conferring resistance to bedaquiline were identified through whole genome sequencing of resistant M.

tuberculosis isolates selected in vitro and were all concentrated in the atpE gene. This gene codes for the c subunit of the transmembrane component (called F0) of the ATP synthase.¹³⁶ Bedaquiline binds the c subunit at the level of amino acids 28, 59, 61, 63, and 66:¹³⁷ Docking studies have shown that bedaquiline has high affinity for the ATP synthase even at low proton motive force and with low pH values.¹³⁸ The binding of bedaquiline to the ATP synthase results in the disruption of the normal functioning of the chain of protonic transfer.¹³⁹ This

24 alteration of the subunit c interface of the F0 component of the ATP synthase leads to futile proton cycling and consequently to death of mycobacteria.¹⁴⁰ In addition, it has been shown that bedaquiline also binds to the ε subunit of the F0 component of the ATP synthase.¹⁴¹ This finding has been further confirmed by showing that the replacement of an amino acid of the chain of ε subunit resulted in bacterial hypersusceptibility to bedaquiline.¹⁴²

Pharmacokinetics

In the mouse model, the oral administration of bedaquiline leads to a prolonged diffusion of the molecule in the tissues and to a long half-life in both plasma (range: 43.7-64.0 h) and tissues (range: 28.1-92.0 h).¹³¹ In humans, the half-life is of approximately 24 h and the terminal half-life is extremely long, up to 4-5 months: this is most likely due to redistribution from the tissue compartments.¹⁴³ The bioavailability of bedaquiline after oral administration is high; the administration with food (standard meal) results in a significant increase in absorption compared with fasting conditions. The peak of plasmatic concentration (CMAX) is reached after a median of 5 h following drug administration. Both CMAX and the area under the curve (AUC) of plasmatic concentrations increase proportionally to the increase of the administered dose of bedaquiline.¹³¹ Overall, bedaquiline activity seems to be principally linked to the AUC and therefore to the cumulative weekly dose, rather than to the frequency of administration: the pharmacokinetic profile seems therefore favourable for intermittent administration of the drug.¹⁴⁴

In a Phase II clinical trial evaluating the early bactericidal activity (EBA) of 400 mg daily bedaquiline, the CMAX reached 5.5 mg/L and the 24-hour AUC was 64,75 mg/h/L.¹⁴⁵ Bedaquiline is metabolised by the liver to a N-desmethyl derivate (M2) that has a residual bactericidal activity, although 5 times weaker than bedaquiline, and then further metabolised to two derivates that have almost no residual activity (M3 and M4). Both bedaquiline and its metabolites have very high levels (>99.9%) of binding with plasmatic proteins. Metabolism is mainly mediated by the CYP3A4 isoenzyme of P450 cytochrome, and less importantly by CYP2C8 and CYP2C19.¹⁴⁶ As a consequence, bedaquiline should be used with caution in patients with mild or severe hepatic impairment; conversely, the drug can be used safely in patients with renal impairment. Elimination is mainly through faeces: the renal clearance is negligible.

Bedaquiline is a lipophilic compound, with good penetration into body tissues such as the lungs. In a study on the C3HeB/FeJ mouse model, which has the advantage of closely mimic the pathophysiological conditions of human lungs as it develops caseous necrotic granulomas, bedaquiline was shown to accumulate more in the highly cellular regions of the lungs, and at a lower but still biologically relevant concentration within the central caseum.¹⁴⁷ Moreover, in mice, it was shown with liquid chromatography / tandem mass spectrometry that bedaquiline administration led to wide distribution into the brain, reaching the CMAX in 4

26 hours after administration.¹⁴⁸ In humans, a similar technique combining liquid chromatography with tandem mass spectrometry has been validated for the measurements of serum levels of bedaquiline.¹⁴⁹ In a case of a MDR-TB patient with tuberculous meningitis who was receiving a bedaquiline-containing regimen, this method showed therapeutic bedaquiline concentrations in the serum, but undetectable levels in the cerebrospinal fluid.¹⁵⁰ This finding has however been criticised with regard to the collection methods that were used and that may not be the most suitable for a highly-lipophilic compound such as bedaquiline.¹⁵¹

Table 1. Drug-drug interactions of bedaquiline and cytochrome P450 inducers/inhibitors.

Modified from the endTB Clinical and programmatic guide for patient management with new TB drugs, version 4.0.¹⁵²

Interaction with

cytochrome P450 Drugs Effect and recommendation

Strong/moderate inducers

Drug-drug interactions

The P450-mediated liver metabolism makes bedaquiline prone to drug-drug interactions with other molecules that inhibit or induce the P450 cytochrome activity. The interaction of bedaquiline with rifamycins have been relatively well studied. The co-administration of rifampicin was associated in the mouse model to a 50% reduction of plasmatic concentrations of bedaquiline, although this did not affect the treatment efficacy.¹⁵³ These findings were confirmed in healthy volunteers, where rifampin reduced bedaquiline AUC by approximately 45%. In the same study, rifabutin, which is also part of the rifamycin family but usually less prone to drug interactions, resulted in little quantitative impact on bedaquiline exposure but was associated with severe (Grade 3 and 4) adverse events before and after the end of the 29-day bedaquiline treatment course.¹⁵⁴ Consequently, the administration of any rifamycin and bedaquiline is contraindicated. More generally, the co-administration of bedaquiline with any moderate or strong inducer or inhibitor of the cytochrome P450 is likely to have relevant pharmacokinetic consequences and should be avoided if possible (see Table 1).

Table 2. Drug-drug interactions of bedaquiline and antiretrovirals.

Modified from the endTB Clinical and programmatic guide for patient management with new TB drugs, version 4.0.¹⁵²

28 Due to the well-known abysmal outcome of MDR-TB treatment in HIV-infected individuals, and the consequent need for a highly-effective treatment in this population, the interactions between bedaquiline and antiretrovirals are particularly relevant. In a Phase I clinical trial, the administration of efavirenz reduced the AUC of bedaquiline by a median of about 20%, which is unlikely to have clinical implications.¹⁵⁵ However, a more recent modelling study predicted efavirenz to reduce average steady-state concentrations of bedaquiline and M2 by 52%.¹⁵⁶ In a pharmacokinetic study, patients receiving a bedaquiline-containing regimen for MDR-TB were divided in three groups: negative patients, HIV-positive patients receiving antiretroviral treatment including nevirapine, and HIV-HIV-positive patients receiving antiretroviral treatment including lopinavir/ritonavir. While there was no difference between the nevirapine group and HIV-negative group, bedaquiline blood levels were significantly increased by lopinavir/ritonavir. No effect was reported on any treatment group on metabolite M2 exposure.¹⁵⁷ Using a previously developed pharmacokinetic model, the authors subsequently estimated lopinavir/ritonavir to reduce bedaquiline clearance to 25%

(17–35%) and M2 clearance to 59% (44–69%) of original values. Conversely, nevirapine modified bedaquiline clearance to 82% (95% CI 67–99%) and M2 clearance to 119% (92–

156%) of original values, likely not leading to any clinically-significant interaction.¹⁵⁸ Overall effects of co-administration of antiretrovirals on bedaquiline plasmatic concentrations and treatment recommendations are summarised in Table 2.¹⁵²

The pharmacokinetic interaction between bedaquiline and other second-line anti-tuberculosis has not been studied into detail, except for the co-administration of clofazimine that was not found to lead to a statistically significant effect on bedaquiline concentrations.¹⁵⁹

Pre-clinical studies: in vitro

The spectrum of activity of bedaquiline is almost completely restricted to mycobacteria. In the first study describing the in vitro activity of this drug, the minimum inhibitory concentrations (MIC) of bedaquiline on M. tuberculosis were between 0.03 and 0.12 mg/L. As expected, in light of the novel mechanism of action compared to the other anti-tuberculosis compounds, the same MICs were reported on MDR M. tuberculosis strains.¹³¹ Another studies analysed a significant number of drug-susceptible and MDR M. tuberculosis strains and identified the median MIC at 0.03 mg/L.¹⁶⁰ Remarkably, bedaquiline was shown to be equally effective in vitro against dormant mycobacteria as against actively replicating ones.¹³⁶ This finding was replicated in a dormancy model where bedaquiline and other lipophilic drugs were more active than hydrophilic agents against dormant M. tuberculosis strains in hypoxic conditions at low pH.¹⁶¹ The in vitro bactericidal action mediated by bedaquiline seems to be more effective on intracellular mycobacteria.¹⁶²

Several studies have assessed the synergy of bedaquiline with other anti-mycobacterials and with other drugs. In whole blood cultures of healthy volunteers, the combination of rifabutin plus bedaquiline produced enhanced intracellular mycobactericidal activity compared the sum of their individual effects. This synergy not found between bedaquiline and rifampicin.¹⁶³ Another study on whole blood cultures testing different drug associations against M. tuberculosis suggested that the most active drug combinations were those containing bedaquiline plus sutezolid (a component of the oxazolidinone family, like linezolid) or SQ109 (a ethylenediamine derived from ethambutol), two promising compounds currently being developed, which were predicted to have cumulative activity comparable to standard first-line anti-tuberculosis treatment. In this study, pretomanid (a nitroimidazole like delamanid) and bedaquiline interacted antagonistically.¹⁶⁴ The co-administration of SQ109 increases the bacterial killing and post-antibiotic effect of bedaquiline, and reduces its MIC.¹⁶⁵ Similar findings have been reported for the association of bedaquiline and BTZ043 (a benzothiazinone).¹⁶⁶ The combination of linezolid and bedaquiline has also shown an additive bactericidal activity against both dormant and actively replicating populations.¹⁶⁷

In a recent study, the administration of verapamil, a calcium channel blocker and an inhibitor of mycobacterial efflux pumps, was shown to potentiate the in vitro activity of bedaquiline. However, this was not the consequence of increased intrabacterial bedaquiline, as expected, but rather the result of verapamil’s ability to directly impact membrane

30 energetics¹⁶⁸ and to enhance systemic exposure to companion drugs via effects on mammalian transporters.¹⁶⁹

Bedaquiline appears as a promising candidate as part of treatment regimens for other non-tuberculous mycobacteria, such as M. ulcerans, M. abscessus, M. fortuitum, M. kansasii, M. marinum, and the mycobacteria belonging to the M. avium complex (MAC).131,160,170 Other species, like M. xenopi, appear to be naturally resistant to this drug.¹⁷¹ However, a detailed summary of the efficacy of bedaquiline on non-tuberculous mycobacteria is outside of the scope of this work.

Pre-clinical studies: animal model

There is a rich amount of evidence on bedaquiline activity coming from pre-clinical studies performed in the animal model and in particular in mice. These studies provided the first in vivo data on bedaquiline. In the non-established TB infection mouse model, bedaquiline showed a bactericidal activity on M. tuberculosis starting from a minimum posology of 12.5 mg/kg 4 times weekly. The addition of bedaquiline to the standard first-line anti-tuberculosis regimen led to a significant increase of the bactericidal efficacy.¹³¹ Similar results were obtained by studies testing the activity of bedaquiline in guinea pigs.^172,173

A primary objective of studies in mice is testing the activity of bedaquiline in combination with other antimycobacterials. This has been done using different protocols in the established TB infection mouse model in order to test bactericidal and sterilising activity.

The combination of bedaquiline and pyrazinamide was shown to exert a synergistic bactericidal activity in mice, compared to other combinations including isoniazid, rifampicin, and moxifloxacin.¹⁷⁴ Moreover, this combination has also demonstrated remarkable sterilising capabilities.¹⁷⁵ These results support the findings that combinations including bedaquiline and pyrazinamide might have the potential to shorten the treatment of drug-susceptible tuberculosis from 6 to 4 months.¹⁷⁶ Similarly, once-weekly administration of bedaquiline, in combination with pyrazinamide and rifapentine, appears to have a sterilising efficacy that exceeds the one of the first-line daily anti-tuberculosis regimen.¹⁷⁷ Bedaquiline has also been tested in association with second-line drugs. The combination of bedaquiline plus amikacin, moxifloxacin, pyrazinamide, and ethionamide, allowed to sterilize the culture of mice lungs

31 after only two months of treatment (bactericidal activity)¹⁷⁸ and to prevent the appearance of relapse 6 months after the end of treatment (sterilising activity).¹⁷⁹ Moreover, drug combinations including only new or experimental drugs, such as bedaquiline plus clofazimine, sutezolid, and/or pretomanid, with or without pyrazinamide, were proven to exert both bactericidal and sterilising efficacy on M. tuberculosis in mice, with the potential to surpass the efficacy of current standard first-line treatment.¹⁸⁰ These findings were replicated using the combination of bedaquiline, pretomanid, and linezolid,¹⁸¹ and the association of bedaquiline with sutezolid and a nitroimidazole between pretomanid and TBA-354, with or without pyrazinamide.¹⁸² In a recent study, the combination of bedaquiline plus pretomanid, moxifloxacin and pyrazinamide, rendered all mice relapse-free after only two months of treatment. In addition, data provided by this study seemed to suggest that pyrazinamide could be discontinued after the first month of treatment without compromising the sterilising activity of the drug combination.¹⁸³

Among non-antimycobacterials, bedaquiline has been tested in combination with verapamil in two studies, showing that the coadministration of verapamil with subinhibitory doses of bedaquiline gave the same bactericidal effect in mice as did the full human bioequivalent bedaquiline dosing.^169,184 In contrast with what was expected, the administration of verapamil had no effect on clofazimine activity.¹⁶⁹

One study has assessed in mice the interest of bedaquiline in the treatment of latent tuberculosis infection: a monotherapy with bedaquiline for 3 to 4 months may be a promising alternative to current standard treatment.¹⁸⁵

Among non-tuberculous mycobacteria, studies in mice have shown that bedaquiline has bactericidal activity against M. leprae¹⁸⁶ and M. ulcerans,¹⁷⁰ bacteriostatic activity against M. avium,¹⁸⁷ and almost no activity on M. abscessus.¹⁸⁸

Figure 11. Seven-days early bactericidal activity of bedaquiline.

Seven-day early bactericidal activity of bedaquiline (TMC207 in the figure) at different doses between 25 and 400 mg daily, compared to rifampin and isoniazid. Source: Rustomjee et al.¹⁸⁹

Figure 12. 14-days early bactericidal activity of bedaquiline.

14-days early bactericidal activity of bedaquiline (BDQ in the figure) at different doses between 100 and 400 mg daily, preceded by a loading dose, and compared to the standard anti-tuberculosis treatment (Rifafour in the figure). Source: Diacon et al.¹⁹⁰

Efficacy: clinical trials

High-quality evidence on the efficacy of bedaquiline is unfortunately lacking: so far, results from only four Phase IIa and three Phase IIb clinical trials are available. The manufacturer engaged to perform a Phase III clinical trial when receiving the provisional approval by the United States Food and Drug Administration (FDA) in 2012: however, the STREAM 2 trial (ClinicalTrials.gov identifier: NCT02409290), a Phase III randomized controlled clinical trial including two bedaquiline-containing arms,¹⁹¹ has only started recruiting in 2017 and results are not expected before late 2021.

The Phase IIa studies assessed the EBA of bedaquiline and other drugs or drug combinations. EBA studies are performed by collecting sputum at baseline and during each day of treatment, following study drug intake, up to a pre-specified number of days; sputum sample are cultured, and the number of colony forming units (CFU) are counted after a few weeks of incubation at the dilution that yields visible colonies. Theefficacy of treatment arms is then assessed by the change in log10 CFU/mL sputum from baseline.

In two EBA studies, bedaquiline was tested alone against different comparators. In the first study, the 7-day EBA of bedaquiline at different posologies (25 to 400 mg daily) was compared to the one of rifampicin and isoniazid: significant bactericidal activity of 400 mg daily of bedaquiline was observed from day four of administration and was similar in magnitude to comparators over the same period, although lower overall (Figure 11). These findings suggest that bedaquiline exerts a late-onset bactericidal activity with a clear relationship between dose and response.¹⁸⁹ In the second study, the 14-day EBA of standard first-line anti-tuberculosis treatment was compared to the one of bedaquiline given at daily doses of 100 mg, 200 mg, 300 mg, and 400 mg, preceded by loading doses (200 mg, 400 mg, 500 mg, and 700 mg on the first treatment day, and 100 mg, 300 mg, 400 mg, and 500 mg on the second treatment day, respectively). The EBA in the 400 mg dose group was greater than that in the 100 mg dose group, although significantly inferior to the one of the first-line treatment association. All the bedaquiline groups showed significant delayed-onset bactericidal activity that was continued to the end of the 14-day evaluation period (Figure 12).

The results of this dose-ranging study confirmed the existence of a linear dose/efficacy trend and supported the use of a loading dose for bedaquiline.¹⁹⁰

Figure 13. Time to sputum culture conversion of bedaquiline and placebo arm, trial C208 Stage 1.

Kaplan-Mayer curve showing the time to culture conversion of sputum samples of MDR-TB patients receiving either bedaquiline (TMC207) or placebo plus optimised background regimen in study C208 Stage 1. Source: Diacon et al.¹⁹²

33 In the other two EBA studies, bedaquiline was tested as part of multidrug regimens.

One study tested the 14-day EBA of parallel groups receiving 1) bedaquiline, 2) bedaquiline-pyrazinamide, 3) PA-824-bedaquiline-pyrazinamide, 4) bedaquiline-PA-824, 5) PA-824-moxifloxacin-pyrazinamide, and 6) standard antituberculosis treatment. The bedaquiline-containing regimens showed a good EBA, although in average lower compared to PA-824 (pretomanid)-based regimens. Among bedaquiline-containing arms, the highest EBA was exerted by the combination of bedaquiline and pyrazinamide, confirming the findings of existing synergy between these two drugs.¹⁹³ Another, more recent study had a similar approach assessing the 14-days EBA of seven treatment arms: 1) bedaquiline-pyrazinamide-clofazimine; 2) bedaquiline-pretomanid-pyrazinamide; 3) bedaquiline-pretomanid-pyrazinamide-clofazimine;

4) bedaquiline-pretomanid-clofazimine; 5) clofazimine alone; 6) pyrazinamide alone; 7) standard first-line treatment. The highest EBA was recorded in the bedaquiline-pretomanid-pyrazinamide arm; this treatment association showed a significantly higher bactericidal activity than the standard first-line treatment.¹⁹⁴

The results of these EBA studies allowed characterising the bactericidal activity of bedaquiline, defining its posology, and identifying the most promising drug combinations to bring forward in subsequent clinical trials. However, the main sources of evidence, which led to the provisional approval of bedaquiline for MDR-TB treatment, are two Phase IIb studies, C208 Stage 1 and C208 Stage 2. C208 Stage I is a randomized, controlled trial, where 47 patients with pulmonary MDR-TB were randomly assigned to receive either bedaquiline (n=23) (400 mg daily for 2 weeks, followed by 200 mg three times a week for 6 weeks) or placebo (n=24) in combination with an optimised background regimen. The primary efficacy

Dans le document The DART-Europe E-theses Portal (Page 40-69)